Caitlin Kalinowski, head of hardware and robotics at OpenAI, announced her resignation, citing concerns over the company's recent military partnership. Kalinowski was a key developer of Meta's augmented reality glasses Orion and joined OpenAI in November 2024. Her departure is seen as a major setback for the company's hardware vision.

In statements on LinkedIn and X, Kalinowski openly pointed out issues such as "surveillance without judicial oversight" and "lethal autonomous weapons that do not require human confirmation," stating that these matters "deserve more consideration than they are currently receiving." She believes the company has been too hasty in making decisions that touch on ethical boundaries.

This personnel upheaval stems from an agreement between OpenAI and the U.S. Department of Defense (the Pentagon) that allows the military to access its AI models. Previously, Anthropic had rejected similar deals for security reasons. Although OpenAI claims its contract includes "hard lines" against mass surveillance and autonomous weapons, Kalinowski and some insiders believe the wording regarding "all legitimate uses" is too vague, leaving significant room for interpretation.

Currently, the U.S. government is trying to standardize such broad cooperation terms, making them a common template for all AI companies working with the nation. Kalinowski's departure not only highlights deep divisions within OpenAI regarding its militarization path but also sparks a new round of discussion in Silicon Valley about the ethical boundaries of AI technology.